Obr et al. Breast Cancer Research (2018) 20:138 https://doi.org/10.1186/s13058-018-1063-2 RESEARCHARTICLE Open Access Insulin-like growth factor receptor signaling in breast tumor epithelium protects cells from endoplasmic reticulum stress and regulates the tumor microenvironment Alison E. Obr1, Sushil Kumar2, Yun-Juan Chang3, Joseph J. Bulatowicz1, Betsy J. Barnes4, Raymond B. Birge2, Deborah A. Lazzarino2, Emily Gallagher5, Derek LeRoith5 and Teresa L. Wood1* Abstract Background: Early analyses of human breast cancer identified high expression of the insulin-like growth factor type 1 receptor (IGF-1R) correlated with hormone receptor positive breast cancer and associated with a favorable prognosis, whereas low expression of IGF-1R correlated with triple negative breast cancer (TNBC). We previously demonstrated that the IGF-1R acts as a tumor and metastasis suppressor in the Wnt1 mouse model of TNBC. The mechanisms for how reduced IGF-1R contributes to TNBC phenotypes is unknown. Methods: We analyzed the METABRIC dataset to further stratify IGF-1R expression with patient survival and specific parameters of TNBC. To investigate molecular events associated with the loss of IGF-1R function in breast tumor cells, we inhibited IGF-1R in human cell lines using an IGF-1R blocking antibody and analyzed MMTV-Wnt1- mediated mouse tumors with reduced IGF-1R function through expression of a dominant-negative transgene. Results: Our analysis of the Molecular Taxonomy of Breast Cancer International Consortium (METABRIC) dataset revealed association between low IGF-1R and reduced overall patient survival. IGF-1R expression was inversely correlated with patient survival even within hormone receptor-positive breast cancers, indicating reduced overall patient survival with low IGF-1R was not due simply to low IGF-1R expression within TNBCs. Inhibiting IGF-1R in either mouse or human tumor epithelial cells increased reactive oxygen species (ROS) production and activation of the endoplasmic reticulum stress response. IGF-1R inhibition in tumor epithelial cells elevated interleukin (IL)-6 and C-C motif chemokine ligand 2 (CCL2) expression, which was reversed by ROS scavenging. Moreover, the Wnt1/dnIGF-1R primary tumors displayed a tumor-promoting immune phenotype. The increased CCL2 promoted an influx of CD11b+ monocytes into the primary tumor that also had increased matrix metalloproteinase (MMP)-2, MMP-3, and MMP-9 expression. Increased MMP activity in the tumor stroma was associated with enhanced matrix remodeling and collagen deposition. Further analysis of the METABRIC dataset revealed an increase in IL-6, CCL2, and MMP-9 expression in patients with low IGF-1R, consistent with our mouse tumor model and data in human breast cancer cell lines. Conclusions: Our data support the hypothesis that reduction of IGF-1R function increases cellular stress and cytokine production to promote an aggressive tumor microenvironment through infiltration of immune cells and matrix remodeling. Keywords: IGF-1R, IL-6, CCL2, Breast cancer, Wnt1, Cellular stress, MMP * Correspondence: [email protected] 1Department of Pharmacology, Physiology & Neuroscience, Rutgers-New Jersey Medical School, Cancer Institute of New Jersey, Newark, NJ 07101, USA Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Obr et al. Breast Cancer Research (2018) 20:138 Page 2 of 17 Background queries of the Cancer Genome Atlas (TCGA) database for Triple-negative breast cancers (TNBCs), classified as es- IGF-1R expression identified higher IGF-1R expression in trogen receptor (ER)-negative, progesterone receptor luminal A and luminal B breast tumors and lower expres- (PR)-negative and lacking human epidermal growth factor sion in HER2-like and triple-negative tumors [15]. Taken receptor 2 (HER2) amplification remain the most aggres- together, these data suggest the function of IGF-1R is sive breast tumor subtype, and approximately 50% of dependent on the tumor type and signaling context. TNBCs classify as basal-like [1]. While chemotherapy Several studies have established that IGF signaling is im- treatment for TNBC primary tumors is effective portant for maintaining cellular stress homeostasis such short-term, tumors ultimately recur and frequently that modifications in IGF signaling result in alterations in metastasize [2, 3]. Hyperactivation of the Wnt pathway is stress signaling. Endoplasmic reticulum (EnR) stress is a common in breast carcinomas where it is often activated consequence of increased misfolded proteins and results in the absence of downstream mutations [4], and Wnt1 in the production of reactive oxygen species (ROS) and ul- overexpression in mammary epithelium is sufficient to timately cell death (for reviews, see [16, 17]. form basal-like tumors in mice with low metastatic poten- Reduction-of-function mutations in the IGF signaling tial [5]. pathway in Caenorhabditis elegans result in activation of Signaling through the insulin-like growth factor type 1 the unfolded protein response (UPR) leading to an en- receptor (IGF-1R) is complex, and defining its role in hanced EnR stress response [18]. Furthermore, activation breast tumorigenesis has been controversial. Early stud- of IGF-1 signaling in breast cancer and neuronal cells pro- ies reported that expression of the IGF-1R correlated tects from EnR-stress-induced apoptosis by enhancing with ER expression and predicted a favorable phenotype EnR stress responses to promote cellular adaptability for [6]. Numerous studies have further confirmed cross-talk cell survival maintenance [19, 20]. Moreover, the inhib- between the ER and IGF-1R in breast cancer (for review, ition of IGF signaling in breast cancer cells results in acti- see [7]). Consistent with these data, loss of IGF-1R has vation of EnR stress to induce autophagy and protect been associated with breast tumor progression into a from apoptosis [21]. These results suggest the IGF path- more undifferentiated phenotype [8], suggesting that way protects cells from EnR stress, and that perturbation IGF-1R is involved in tumor suppression. However, of the IGF pathway leads to enhanced overall EnR stress. other findings have shown that IGF signaling is a posi- In the present study, we tested the hypothesis that at- tive mediator of breast cancer growth and survival (for tenuated IGF-1R function promotes tumor epithelial cell reviews, see [9, 10]). Since IGF signaling promotes stress resulting in tumor stromal environment alter- tumor cell proliferation and survival, various inhibitors ations to establish an aggressive phenotype in breast tu- have been developed to attenuate IGF signaling (for re- mors. We determined that IGF-1R is essential in tumor views, see [11, 12]). Despite early preclinical findings, suppression in breast tumorigenesis. We demonstrate the usefulness of disrupting IGF-1R signaling in clinical that attenuated IGF-1R signaling in the MMTV-Wnt1 trials has been less than promising and, in some cases, mouse mammary tumor model and in human breast inhibiting the pathway has led to worse outcomes [11, cancer cell lines increases tumor epithelial cellular stress, 12]. Collectively, these diverse results support the possi- resulting in upregulation of cytokine production. These bility that the IGF-1R has a dual function as both a changes result in altered migration and infiltration of tumor suppressor and an oncogene. tumor immune cells and dramatic alterations in the In our recent studies, we established a mouse line with tumor microenvironment associated with promoting pri- transgenic expression of a kinase-dead, dominant-negative mary tumor epithelial cell extravasation. IGF-1R (MMTV-dnIGF-1R) in combination with MMTV-Wnt1 expression to test how decreased IGF-1R Methods signaling in the mammary epithelium impacts a Antibodies and reagents well-established mouse model of basal-like breast cancer Rabbit monoclonal anti-phospho-eIF2a (D9G8), rabbit [5]. Attenuation of IGF-1R in this model resulted in de- monoclonal anti-eukaryotic initiation factor 2-alpha creased tumor latency, an enhanced basal phenotype, and (eIF2a) (D7D3), rabbit monoclonal anti-protein disulfide potentiation of lung metastases (Additional file 1:Table isomerase (PDI) (C81H6), mouse monoclonal anti-C/EBP S1, see also [1]). These results were surprising given that homologous protein (CHOP) (L63F7), rabbit monoclonal the MMTV-Wnt1 tumors have low metastatic potential anti-phospho-Akt (Ser473) (D9E), rabbit monoclonal [5]. However, similar findings were reported from condi- anti-Akt (11E7), rabbit monoclonal anti-phospho-IGF-1R/ tionally deleting IGF-1R in a prostate cancer mouse model IR (D6D5L), and rabbit monoclonal anti-IGF-1R (D23H3) [13]. These data are also consistent with new reports that antibodies were purchased from Cell Signaling. Rabbit have correlated high IGF-1R and ERα expression in lu- polyclonal anti-matrix metalloproteinase (MMP)-2 minal B breast